研究深度学习在电子信息材料声参数检测中的应用,通过多传感器阵列采集声信号,并利用卷积神经网络进行参数反演,通过构建多层神经网络实现声学信号的高效处理与特征提取.实验结果显示,应用该技术,可在提高检测精度的同时显著缩短检测时间,相对误差控制在 2%以内,单次检测时间缩短至 2 s.
Acoustic Parameter Detection Technology of Electronic Information Materials Based on Deep Learning
To investigate the application of deep learning in the detection of acoustic parameters of electronic information materials,sound signals are collected through multi-sensor arrays and parameters are retrieved using convolutional neural network,a multi-layer neural network is constructed to realize efficient acoustic signal processing and feature extraction.The experimental results show that the technique can improve the detection accuracy and significantly reduce the detection time,the relative error is controlled within 2%,and the single detection time is reduced to 2 s.
deep learningelectronic information materialsacoustic parameter detection
NETL
NSTL
万方数据
